Deep Learning for Voltammetric Sensing in a Living Animal Brain

Abstract

Numerous neurochemicals have been implicated in the modulation of brain function, making them appealing analytes for sensors and diagnostics. However, it is a grand challenge to selectively measure multiple neurochemicals simultaneously in vivo because of their great variations in concentrations, dynamic nature, and composition. Herein, we present a deep learning‐based voltammetric sensing platform for the highly selective and simultaneous analysis of three neurochemicals in a living animal brain. The system features a carbon fiber electrode capable of capturing the mixed dynamics of a neurotransmitter, neuromodulator, and ions. Then a powerful deep neural network is employed to resolve individual chemical and spatial‐temporal information. With this, a single electrochemical measurement reveals an interplaying concentration changes of dopamine, ascorbate, and ions in living rat brain, which is unobtainable with existing analytical methodologies. Our strategy provides a powerful means to expedite research in neuroscience and empower sensing‐aided diagnostic applications.

Highly Stable and Selective Sensing of Hydrogen Sulfide in Living Mouse Brain with NiN4 Single-Atom Catalyst-Based Galvanic Redox Potentiometry

“…Despite its well-known significance for the CNS, H 2 S is much less understood from the perspective on its dynamics triggered in different physiological and pathological conditions. ,,, For this purpose, various in vivo H 2 S sensing methods were motivated, including fluorescence, − two-photon microscopy, ,, magnetic resonance imaging (MRI), , and colorimetric and electrochemical sensing. − Among them, in vivo imaging techniques are well-suited for mapping cross-regional concentration fluctuations and distributions of H 2 S. ,− Electrochemical sensing methods own high spatial and temporal resolution; , thus, they are broadly utilized for real-time measurement of neurochemical dynamics in rapid neural processes. − However, the existing electrochemical H 2 S sensors suffer from sulfur poisoning, electrode passivation, and persistent sensitivity reduction because of the adsorption of H 2 S oxidation products (S n molecules) − on the electrode surface. Therefore, the development of high-performance electrochemical sensors for accurate H 2 S sensing in vivo remains a constant need.…”

Multiplexed Monitoring of Neurochemicals via Electrografting-Enabled Site-Selective Functionalization of Aptamers on Field-Effect Transistors

“…Because of that, coupling FSCV or rapid pulse voltammetry with voltammogram analysis (multivariate penalized regression, partial least squares regression, and deep learning) has been used to distinguish voltammograms from multiple neurochemicals, thereby making it possible for multiplexed analysis. 32,48,49 FSCV, however, fails to distinguish some structurally similar neurotransmitters, such as dopamine and norepinephrine. 50−52 Microdialysis suffers from large temporal resolution (typically several minutes) and low spatial resolution because of the semipermeable membrane and large probe size (150−440 μm in diameter).…”

“…In past decades, with the advances in neuroscience and micro-/nanofabrication, groundbreaking sensors have been developed to target specific brain regions at different scales. − The main techniques for neurotransmitter monitoring include the following several types: (1) nuclear medicine tomographic imaging, such as positron emission tomography (PET); (2) optical sensing techniques, such as surface-enhanced Raman spectroscopy (SERS), , fluorescence, , chemiluminescence, optical fiber biosensing and colorimetry; (3) electrochemical methods, − like fast-scan cyclic voltammetry (FSCV) − and amperometry; (4) mass spectrometry; ,, and (5) microdialysis sampling (typically coupled with mass spectrometry analysis). − While each of these techniques has its pros and cons, ,, it is still a challenge to build a system that can effectively capture the dynamics of neurotransmitter release with a high temporal resolution, cellular scale spatial resolution, superior sensitivity, and selectivity, not to mention empowering the tools with multiplexed monitoring capabilities. Among these methods, FSCV, microdialysis, and genetically encoded fluorescent sensors are three widely used or emerging techniques for neurotransmitter monitoring .…”

“…Specifically, FSCV possesses better temporal resolution (milliseconds) and sensitivity (nanomolar). Because of that, coupling FSCV or rapid pulse voltammetry with voltammogram analysis (multivariate penalized regression, partial least squares regression, and deep learning) has been used to distinguish voltammograms from multiple neurochemicals, thereby making it possible for multiplexed analysis. ,, FSCV, however, fails to distinguish some structurally similar neurotransmitters, such as dopamine and norepinephrine. − Microdialysis suffers from large temporal resolution (typically several minutes) and low spatial resolution because of the semipermeable membrane and large probe size (150–440 μm in diameter). , Recently, genetically encoded fluorescent sensors have been developed to image the dynamics of neurochemical release in vivo, including DA and glutamate. ,− However, the multiplex monitoring with high selectivity and non-overlapping spectra beyond the dual-color imaging of glutamate and DA is challenging . Additionally, the complicated genetic modification process and the requirements of coupling with fiber photometry for neurochemical monitoring limit the practical applications of these genetically encoded sensors …”

Novel, User-Friendly Experimental and Analysis Strategies for Fast Voltammetry: Next Generation FSCAV with Artificial Neural Networks

“…There are a variety of strategies that can be utilized to improve these analysis challenges. − Artificial neural networks (ANNs) are particularly attractive due to their ability to learn from big data sets, their capabilities to fit nonlinear data, and high accuracy of predictions. ANNs are machine learning models that resemble biological neural networks.…”